Influence of Social Media Marketing, PDF [PDF]

Zitiervorschau

INFLUENCE OF SOCIAL MEDIA MARKETING, BRAND LOYALTY AND eWOM TOWARDS CONSUMERS’ PURCHASE INTENTION

BONG WAN ZHUNG

MASTER OF BUSINESS ADMINISTRATION

UNIVERSITI TUNKU ABDUL RAHMAN FACULTY OF ACCOUNTANCY AND MANAGEMENT DECEMBER 2017

Influence of Social Media Marketing, Brand Loyalty and eWOM towards Consumers’ Purchase Intention

Bong Wan Zhung

A research project submitted in partial fulfillment of the requirement for the degree of Master of Business Administration

Universiti Tunku Abdul Rahman Faculty of Accountancy and Management December 2017

Influence of Social Media Marketing, Brand Loyalty and eWOM towards Consumers’ Purchase Intention

By Bong Wan Zhung

This research project is supervised by:

Dr. Hen Kai Wah Assistant Professor Department of International Business Faculty of Accountancy and Management

Copyright @ 2017 ALL RIGHTS RESERVED. No part of this paper may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, graphic, electronic, mechanical, photocopying, recording, scanning, or otherwise, without the prior consent of the authors.

i

DECLARATION

I hereby declare that: (1) This Research Project is the end result of my own work and that due acknowledgement has been given in the references to all sources of information be they printed, electronic, or personal. (2) No portion of this research project has been submitted in support of any application for any other degree or qualification of this or any other university, or other institutes of learning. (3) The word count of this research report is 20404.

Name of Student:

BONG WAN ZHUNG

Student ID:

13UKM06827

Signature:

Bong Wan Zhung

Date:

30th November 2017

ii

ACKNOWLEDGEMENT

The journey as an MBA student was a thought-provoking and exciting experience for me. Over these few years, there are many people who were directly and indirectly helping and supporting me to achieve my journey. Firstly, I would like to express my deepest gratitude to my family on this four-year journey, and especially my wife Joanne for her unconditional supports. Second, I wish to express my gratitude to Dr. Hen Kai Wah for his support, advice, and guidance in completing this research project. Thirdly, the members of Faculty and Institute of Postgraduate Studies and Research (IPSR) for the knowledge and helpful throughout my study at Universiti Tunku Abdul Rahman. Finally, to my friends, classmates, and others that being very helpful in responding to my questionnaires.

Thank you.

Your sincerely, Bong Wan Zhung

iii

TABLE OF CONTENTS Copyright Page……………………………………...………………….... Declaration................................................................................................. Acknowledgement ..................................................................................... Table of Contents ...................................................................................... List of Tables ............................................................................................. List of Figures ........................................................................................... Abstract ..................................................................................................... CHAPTER 1 INTRODUCTION ............................................... 1.1 Background .......................................................... 1.2 Problem Statement ............................................... 1.3 Research Objectives ............................................. 1.3.1 General Objective ................................................ 1.3.2 Specific Objectives .............................................. 1.4 Research Questions .............................................. 1.5 Significant of Studies ........................................... 1.6 Term of References .............................................. 1.7 Overview of the Report ........................................ 1.8 Summary .............................................................. CHAPTER 2 LITERATURE REVIEW .................................... 2.1 Introduction .......................................................... 2.2 Social Media and Social Media Marketing .......... 2.2.1 Social Media ........................................................ 2.2.2 Social Media Marketing ....................................... 2.3 Brand Loyalty ...................................................... 2.4 Word of Mouth and Electronic Word of Mouth .. 2.4.1 Word of Mouth .................................................... 2.4.2 Electronic Word of Mouth ................................... 2.4.3 Differences between WOM & eWOM ................ 2.5 Purchase Intention ................................................ 2.6 Previous Empirical Studies .................................. 2.6.1 Social Media Marketing to Brand Loyalty .......... 2.6.2 Electronic Word of Mouth to Brand Loyalty ....... 2.6.3 Social Media Marketing to Purchase Intention .... 2.6.4 Brand Loyalty to Purchase Intention ................... 2.6.5 Electronic Word of Mouth to Purchase Intention. 2.7 Theoretical Framework ........................................ 2.7.1 Theory of Reasoned Action ................................. 2.7.2 Theory of Planned Behaviour .............................. 2.8 Conceptual Framework ........................................ 2.9 Hypotheses Developments ................................... 2.9.1 Social Media Marketing and Brand Loyalty ........ iv

Page i ii iii iv vii viii ix 1 2 4 6 7 7 8 8 10 11 13 14 14 15 15 16 18 20 20 21 22 23 25 26 26 27 27 28 28 28 30 32 34 34

2.9.2 2.9.3 2.9.4 2.9.5

CHAPTER

CHAPTER

CHAPTER

2.10 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.7.1 3.7.2 3.7.3 3.8 3.9 3.10 3.10.1 3.10.2 3.10.3 3.11 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 5 5.1 5.2 5.3 5.3.1 5.3.2

5.3.3

5.3.4 5.3.5

Electronic Word of Mouth and Brand Loyalty .... Social Media Marketing and Purchase Intention.. Brand Loyalty and Purchase Intention ................. Electronic Word of Mouth and Purchase Intention ............................................................... Summary .............................................................. METHODOLOGY .............................................. Introduction .......................................................... Research Design .................................................. Research Approach .............................................. Research Methodological Choices ....................... Population and Sampling ..................................... Sampling Size ...................................................... Data Collection Methods / Instruments ............... Pilot Study ............................................................ Construct Measurement ....................................... Data Scale of Measurement ................................. Variables .............................................................. Reliability Testing ................................................ Statistical Testing ................................................. Linear Regression ................................................ Multiple Regression Analysis .............................. Mediation Effect Analysis ................................... Summary .............................................................. RESULTS ............................................................ Introduction .......................................................... Respondent Demographics .................................. Descriptive Analyses ........................................... Reliability Analysis .............................................. Linear Regression Analysis ................................. Multiple Regression Analysis .............................. Mediation Effect Analysis ................................... Summary .............................................................. DISCUSSION ...................................................... Introduction .......................................................... Findings through Descriptive Analysis ................ Discussion of Major Findings .............................. H1: Social media marketing will have significant positive influence on consumers’ brand loyalty... H2: Electronic word of mouth will have significant positive influence on consumers’ brand loyalty ........................................................ H3: Social media marketing will have significant positive influence on consumers’ purchase intention ............................................................... H4: Brand loyalty will have significant positive influence on consumers’ purchase intention ........ H5: Electronic word of mouth will have significant positive influence on consumers’ purchase intention ................................................ v

35 36 37 37 38 39 39 40 43 45 47 49 50 51 52 54 55 55 57 57 57 58 58 59 59 60 65 67 69 74 76 78 80 80 81 82 83 83

84

84 85

References Appendix

5.3.6 Mediating Effects ................................................. 5.4 Implications ......................................................... 5.5 Limitations of Study ............................................ 5.6 Recommendations for Future Research ............... 5.7 Summary .............................................................. ............................................................................................. ............................................................................................. Appendix A ..................................................................... Appendix B ..................................................................... Appendix C .....................................................................

vi

85 86 87 88 89 91 101 101 106 108

List of Tables Table 1:

Construct Measurements

Page 52

Table 2:

Cronbach’s Alpha Coefficient Range

56

Table 3:

Respondents Demographic Information

60

Table 4:

Reliability Analysis

68

Table 5:

Model Summary (H3)

69

Table 6:

ANOVA (H3)

69

Table 7:

Coefficients (H3)

70

Table 8:

Model Summary (H4)

71

Table 9:

ANOVA (H4)

71

Table 10:

Coefficients (H4)

71

Table 11:

Model Summary (H5)

72

Table 12:

ANOVA (H5)

73

Table 13:

Coefficients (H5)

73

Table 14:

Model Summary (H1, H2)

74

Table 15:

ANOVA (H1, H2)

75

Table 16:

Coefficients (H1, H2)

75

Table 17:

Hypotheses Results

79

vii

List of Figures Figure 1:

Theory of Reasoned Action

Page 29

Figure 2:

Theory of Planned Behaviour

32

Figure 3:

Conceptual Framework

34

Figure 4:

Gender Demographic Information

61

Figure 5:

Age Demographic Information

62

Figure 6:

Occupations Demographic Information

63

Figure 7:

Years of Internet Utilisation

64

Figure 8:

Social Media Awareness

64

Figure 9:

Popular Social Media Platforms in Malaysia

65

Figure 10:

Active Age Groups on Social Media Platform

66

Figure 11:

Brand Consciousness Among Gender

67

Figure 12:

Mediation Effect (1)

77

Figure 13:

Mediation Effect (2)

78

Figure 14:

Conceptual Framework

79

viii

Abstract

Influence of Social Media Marketing, Brand Loyalty and EWOM towards Consumers’ Purchase Intention

Consumers’ purchase intention has evolved with the facilitation of the social media marketing, and gradually replacing the traditional marketing strategies. The marketing strategy is built on social media platforms where consumers are actively seeking information about a product or service, and others experiences before committing to buy the product or service. The sharing of the brands, product reviews, and users’ experiences are part of the electronic word of mouth where information is transmitted and received over the internet. With the influence of electronic word of mouth and social media marketing, it has an impact to the brand loyalty, and eventually towards purchase intentions. This study attempts to study the impact of social media marketing, brand loyalty, and electronic word of mouth towards purchase intentions. Two hundred questionnaires were collected at the end of the survey within Malaysia. Five hypotheses were developed and tested. The results indicated that the social media marketing, brand loyalty, and electronic word of mouth are effectively influencing consumers’ behaviour leading to purchase or repurchase intentions. These findings indicate social media marketing managers need to emphasise the importance of positive word of mouth in leaving a positive impression on consumers and increase engagement with consumers in building brand loyalty.

ix

CHAPTER 1

INTRODUCTION

This chapter proposes a research that will add to an understanding of the consumers’ purchase intention toward products or services in Malaysia. In this research, it will specifically investigate the factors of social media marketing, brand loyalty, and electronic word of mouth (eWOM) in influencing the consumers behavioural towards purchase intentions.

This chapter is organised by discussing the

background of the research, the problem statement, research questions, research objectives, term of references, an overview of the report outline, and finished with the summary of the chapter.

Page 1 of 108

1.1 Background

Ever since the introduction of World Wide Web 28 years ago (World Wide Web Foundation, n.d.), the development of the information age has changed how people used to interact in the past. For the first time in history, the internet created a single social and cultural setting for everyone and as well evolved on how information shares among each other, the way human conducts businesses, and the transformation of technologies.

The development of social media on the internet has changed the way people interact with each other and sharing of information. Social media is well accepted and popular among the millennial cohorts (Nanji, 2017). The millennial is the emergence of a new generation that is powerful and sophisticated consumers who are difficult to influence, persuade, and keep (Devi & Menezes, 2016). This group of cohorts is heavily active on the social media networking sites and always hungry for information. The entire marketing landscape of the traditional marketing is shifting towards social media marketing (Erdoğmuş & Çiçek, 2012). Businesses are now more active on social media compare to the past (Hainla, 2017). Some businesses have more than one presences on the social media platforms to interact with the social media users. It is not necessary to be a presence on the social media to build the brand awareness, but without it, they are very likely to be far behind from their competitors that are actively engage with their fans.

Social media marketing offering businesses an interactive platform to interact with consumers in building the brands, and loyalty that beyond what traditional methods can offer in the past (Erdoğmuş & Çiçek, 2012). Consumers are using the social media as the source to look for product information such as the brand, manufacturer background, and retailers when deciding to buy it (Ler, 2014). The purchase decisions are highly influenced by the electronic word of mouth on the social media sites. The persuasion effect of the electronic word of mouth is seen as an opportunity for businesses to influence the consumers by using various methods to influence Page 2 of 108

them on purchase decisions. Consumers are looking beyond pricing for the product when they are researching for it over the internet. It is also the experience and expectation that Malaysian consumers want (Asean Up, 2017).

The 2016 Nielsen’s surveys on Global Retail Growth Strategies and Global Retail Loyalty Programmes concluded that three in five Malaysian consumers enjoy taking the time to find bargains (Nielsen, 2016). Asean Up (2017) pointed out that Malaysian e-commerce market will continue to flourish as consumers are willing to shop for more new products categories, satisfaction that beyond pricing such as the ease of online payment methods, and express delivery for the products. The allegiance to brands is very important for businesses to sustain the fan base and sustainability. However, it may not be the same for consumers especially the millennial cohorts when they are spoilt for choices. Advantageous promotions and low pricing may motivate the change of brand’s allegiance and it takes a very long effort for businesses to win back the consumer’s heart when they are many related products in the market. In a report by Monteiro (2016), 3 out of 4 online shoppers globally abandon their shopping carts before checking out. He pointed out that when consumers have more options, the expectations become higher, particularly how they engage with the business and receive their purchase. Hence, the highly attractive price is not only the factors but strong customer service and delivery options too. In this regard, consumers are actively sharing their experiences with the retailer on social media sites and eventually will influence loyal customers to think twice before checking out. Thus, not only social media marketing is important, but good electronic word of mouth is important as well to build brand loyalty for the businesses.

A report by We Are Social and Hootsuite (Kemp, 2017), in January 2017 ranks Malaysia in the fifth place globally with the most active social media accounts. A significant increase from January 2016 from 18 million active social media users to 22 million in 2017 (Kemp, 2017). Hamid (2016) presented that 32% of Malaysian population are social media addicts. In general, they spend at least 16 hours a week on social media platform to consume for information and as well actively generating Page 3 of 108

new contents and opinions. And out of the 32% of social media addicts, about 63% are aged below 35 years old.

Malaysia e-commerce is aiming to be the next e-commerce boom in South East Asia according to Brewer (2017) viewpoint. The country has over 136% of mobile penetration rate in 2016 and 47% of them using their smartphones to shop online. And about 7 million consumers shop online every month buying daily necessities, fashion, and accessories, and unusual or hard to find items. Besides that, 30% of all e-commerce transactions in South East Asia are social sales where consumers purchase over the social media sites such as Instagram or Facebook. In fact, Alibaba group is building their regional distribution hub in Malaysia for Malaysia and South East Asian region to boost the region e-commerce activities (EcommerceIQ, 2017).

All these facts are important hints for businesses to acknowledge the importance of social media marketing in pushing marketing information on the social media platform that they do not want to miss it. The one-way traditional marketing methods may not be that useful in this digital world because of the social interactivities on social media sites. A business successfully sets up their presence on social media is not the end, but just the beginning of this new ecosystem. Businesses need to constantly engage on the social media with postings and reviews that capture the heart of the users, and eventually has the intention to buy their products.

1.2 Problem Statement

Until the recent years, the e-commerce activities in Malaysia were slow, according to Brewer (2017). With the keen commitment by the Malaysian government to grow the e-commerce sector, and the regional distribution hub set up by Alibaba Group, Malaysia would be the centre of e-commerce in the upcoming years in South East Page 4 of 108

Asian region. E-commerce merchants in Asia Pacific were facing extremely high competitions among each other as everyone is trying to acquire a piece of the market shares (Monteiro, 2016). The vast options available for consumers, consumers will continuously seek for the best offers before committing to the purchase, and best offer is not necessarily about the pricing only. It is as well the engagement with the brands, companies, and how they receive their purchase. It is all about engagement between businesses and consumers on the internet. In other words, consumers look beyond pricing, such as how business engage with them through social media, the allegiance to the brand through marketing and word of mouth efforts, and how this action leads to the desire to commit a purchase.

Consumers are progressively shifting away from traditional media towards the digital media for information. This is a trend where marketers are less emphasising on traditional marketing channels. A digital content company such as Rev Asia, the largest digital media entity in Malaysia and ranks third after Facebook, and Google in Malaysia based on traffics was acquired by traditional media company Media Prima recently (Russell, 2017; The Star, 2017). It shows that digital content on the internet, especially on social media platform, is rapidly replacing the importance of traditional media especially in Malaysia when the country is top five active on social media globally. Consumers are actively engaging on social media platform to contribute and as well seeking for information about a product, brand before deciding to buy or not. There are high chances that companies in Malaysia will be outpaced by established companies that are actively using social media as the marketing agent to promote their brands, products, or services.

Brand loyalty would be experiencing a notable change as marketers no longer in the position of the hierarchical authority in pushing brand’s awareness that happened in the past with traditional marketing channels. Businesses and consumers are co-creating together the brand loyalty through various engagement activities on social media platform. It is important for marketers or businesses to understand the marketing efforts on social media platform and as well building brand loyalty as consumers may consider liking a brand page on Facebook is Page 5 of 108

considered loyal to the brand. In other words, brand loyalty in the eyes of millennials may be different compared to earlier generations.

The traditional word of mouth happens from person to person communication, and normally involving people they trusted such as family members, friends, peers, and so on. However, the social media platform has become the novel word of mouth platform. Now, consumers are actively sharing their experiences of a brand or products that they used before to the mass on social media platform. This sharing is similar to word of mouth which is non-commercial driven. However, on social media platform, the trusted source of word of mouth has changed from a known person to anonymous person in the online community. And businesses as well can drive the electronic word of mouth on the internet too. Thus, the change of word of mouth to electronic word of mouth seeks a lot of understanding for the marketers and businesses in building a positive brand experience towards the consumers.

On the existing literature reviews, there are many research had been conducted in the recent years in overseas on the social media marketing, electronic word of mouth, and brand loyalty. While in Malaysia, there are research done in recent years on other dimensions influencing towards consumers’ purchase intentions but are limited (Azira, Siti, Law, Nurliyana, & Siti, 2016; Chong, 2014; Balakrishnan, Dahnil, & Wong, 2014; Lim, Osman, Salahuddin, Romle, & Abdullah, 2016; Severi, Kwek, & Nasermoadeli, 2014). Hence, it is important to further validate the Malaysian consumers purchase behaviour.

1.3 Research Objectives

Based on the stated problem statement in the previous section, this section will begin with the research’s general objective. Subsequently, a breakdown of the specific objectives to examine the purpose of this research. The objectives are presented as following: Page 6 of 108

1.3.1 General Objective

The purpose of this research is to identify and examine consumers’ purchase intention in Malaysia through social media marketing, brand loyalty and electronic word of mouth. Besides that, this research is keen to ascertain the relationships between of both social media marketing and electronic word of mouth towards brand loyalty among the consumers in Malaysia leading to purchase intentions.

1.3.2 Specific Objectives

The following are the specific objectives of this research derived from the general objective in the previous section. RO1: To investigate the association between social media marketing and, brand loyalty. RO2: To considers the relationship between electronic word of mouth towards brand loyalty. RO3: To identify the significant influence of social media marketing towards consumers’ purchase intention. RO4: To examine the correlation between brand loyalty and consumers’ purchase intention. RO5: To investigate the significant impacts of electronic word of mouth by communities’ members towards consumers’ purchase intention.

Page 7 of 108

1.4 Research Questions

Based on the defined objectives of section 1.3.2, there are five research questions the researcher is interested to seek for the answers at the end of the research. The research questions are as follows: RQ1: Is there any committed relationship from social media marketing towards brand loyalty? RQ2: How does electronic word of mouth positively affect consumers’ brand loyalty? RQ3: How does social media marketing influences consumers’ behaviour on purchase intention of a product or service? RQ4: How does brand loyalty affect consumers’ behaviour on purchase intention of a product or service? RQ5: How does electronic word of mouth impact the consumers’ purchase intention of a product or service?

1.5 Significant of Studies

The entire South-East Asia region is seeing a boom in the e-commerce, and Malaysia is targeting to be the centre of the e-commerce in the coming years. Through the government commitments and initiatives, and as well as private sector such as Alibaba Group eyeing on Malaysia as the regional distribution hub for their e-commerce businesses. E-commerce merchants are expecting a fierce fight to be competitive, brand awareness, and the engagement with the consumers to attract and retain customer base.

Page 8 of 108

Therefore, this research played a pivotal role to provide an understanding of the current e-commerce context, especially in Malaysia. The critical elements investigated in this research has significant influence on the consumer’s decision to purchase a product or service. Consumers are progressively shifting away from traditional media to digital media for day-to-day information. Marketers that advocate traditional marketing media now could explore digital marketing on various platforms. Social media marketing does works differently from traditional marketing efforts (Erdoğmuş & Çiçek, 2012). It is a platform where both business and consumers co-creating and receiving information about a product or service. The information created through social media marketing played a significant role in influencing other consumers in making purchase decisions.

The social media marketing stands as a medium to advocate marketing to the consumers. However, electronic word of mouth is an important element in social media marketing where positive or negative messages can be forwarded or shared with a greater audience at a fraction of times. It can create significant repercussion where it would help marketers enhance the marketing efforts to better reaching the wider audiences.

The trivial of retaining loyal customers are ever tougher with digital marketing channels. Consumers are exposed to many brands every day on social media platform and there is a notable change in the brand loyalty. The evolving needs of different generations in this digital age, and the loyalty pledge to brands are no longer satisfactory because consumers are widely exposed to newer brands every day on social media that are thriving to establish the connection with consumers rather promoting the products. Therefore, it is an opportunity for marketers to effectively engage Malaysian consumers to build the brand affections.

Page 9 of 108

1.6 Term of References

The terms used in this research are widely defined and interpreted based on different contexts. Hence, it is necessary to provide clear definitions of the key terminologies used in the present research. The following terminologies are defined in the context of this research and presented as follows:

Social Media – Social media translated as the “websites and applications that enable users to create and share content or to participate in social networking” (Oxford University Press, n.d.). Example of social media: Facebook, Instagram, Twitter, LinkedIn, Pinterest, etc.

Social Media Marketing – Social Media Marketing described as a form of Internet marketing utilised the social media as a marketing tool where users produce and share content in helping business to increase brand experience and widen customer reach (Rouse, n.d.) that is not possible via traditional marketing channels (Weinberg, 2009).

Brand Loyalty – Brand loyalty does not result only in constant purchase pattern, it can also bring out positive impression towards the brand, which is not essentially tied with the purchase of the brand (Aaker, 1991; Gounaris & Stathakopoulos, 2004; (Severi et al, 2014).

Word of Mouth (WOM) – An unpaid form of marketing effort to pass experiences worthy information to other people how much they like or dislike a brand, product, or service (Entrepreneur Media Inc., n.d.; Whitler, 2014).

Electronic Word of Mouth (eWOM) – eWOM refers to the impact upon the persuasive power of communities’ members over the social media platform Page 10 of 108

involving the brand, and product (Anderson, Black, Machin, Marcouse, & Watson, 2014; Danieli, 2016; Severi et al, 2014).

Purchase Intention – Purchase intention is well-defined as the possibility the consumer will purchase the products (Danieli, 2016) in the future which reflects their intentions to purchase the products based on their attitudes and emotions (Kim, Chun, & Ko, 2017).

1.7 Overview of the Report

This research attempts to identify the impact of social media marketing, brand loyalty and electronic word of mouth on consumers’ purchase intention, and to assess the relationship among social media marketing, and electronic word of mouth towards brand loyalty.

Chapter One: Introduction

Chapter one, the introduction, provides a broad overview that highlights the rationale behind the research topic. This is done by emphasising the relevance of the research topic and the importance in Malaysia consumers behaviour towards purchase intention of a product or services. Meanwhile, this chapter is structural including the Research Questions, and Objectives, Significant of Studies, Term of References, and concluded with a Summary for this chapter.

Page 11 of 108

Chapter Two: Literature Review

The chapter provides a comprehensive literature review starting with the general introduction of the topic in general and focuses context in Malaysia. Followed by the behavioural studies of consumers that lead to the discussions of each independent variables. The chapter structure is followed by previous empirical studies, the theoretical framework, and the conceptual framework before leading to the hypotheses developments. This chapter is divided into 10 sub-headings aiming at discovering how social media marketing, and electronic word of mouth can affect brand loyalty, and from different aspects to find how social media marketing, brand loyalty, and electronic word of mouth can affect consumers purchase intentions.

Chapter Three: Methodology

Chapter three will outline the steps on how the research is intended to be carried out in terms of the research process design, type of research, population, the variables, the instruments, or methods used for data collection, and various statistical testing methods to support the data analysis in Chapter Four.

Chapter Four: Results

In this chapter, the raw collected data will be housekeeping before feeding into statistical software IBM SPSS Statistic version 22 for detailed statistical analyses. Furthermore, the outcome of the statistical analyses will be presented in this chapter and as well.

Page 12 of 108

Chapter Five: Discussions

This chapter will discuss the finding from the research, interpretation of the findings and whether the hypotheses are supported by the data or not. Furthermore, this chapter will include the discussion of limitations, and recommendations for future research.

1.8 Summary

The development of social media has changed the way people interact with each other and how information is gathered and shared around. This is very common nowadays with the millennial cohorts and progressively impacting the marketing efforts especially promoting the brand, product or service are done over the internet on social media platforms through electronic word of mouth. Through the development of social media marketing as a popular marketing channel over traditional marketing channel, it also influenced the consumers brand allegiance too. Business and consumers are co-creating brand loyalty together through various engagements and different as compared to earlier generations. Hence, the research may provide insightful information how Malaysian consumers react to social media marketing, electronic word of mouth, and brand loyalty towards purchase intention of a product or service.

Page 13 of 108

CHAPTER 2

LITERATURE REVIEW

In this chapter, the literature review is organised with comprehensive reviews of the research topic through secondary information. As the chapter develops, the reviews of relevant theoretical models are discussed before developing the conceptual framework. In the following section will develop the hypotheses based on the conceptual framework before ending with a summary of this chapter.

2.1 Introduction

The use of social media as the next generation marketing tool has changed the way marketers used to market the products and services in traditional marketing channels. Marketers are actively pushing advantageous marketing campaigns to attracts the online community. In response to the competitive social media Page 14 of 108

marketplace, brands are eager to win the heart of the millennials through electronic word of mouth.

Hamid (2016) reported that 32% of Malaysian are social media addicts, spending more than 16 hours online in a week. With the high number of hours online a week, it is an advantage for the brands and marketers to target them on social media. With the online shopping trend is growing, consumers expectation on online shopping will challenge brands and marketers beyond prices (Asean Up, 2017). According to EcommerceIQ (2017), Malaysia e-commerce market was estimated at US$1billion back in 2015 and expected to increase the online shopping market size by eight-folds to US$8billion. Hence, it is important to consider the factors influencing the consumers’ purchase intentions.

2.2 Social Media and Social Media Marketing

2.2.1 Social Media

The term of “Social Media” is constructed from communication science and sociology which explained in the communication context as a means for storing or delivering information while social networks are social structures made up of a set of social actors with a complex set of dyadic ties among them (Peters, Chen, Kaplan, Ognibeni, & Pauwels, 2013). Further elaborate that, social media are 2-way communication channels which enables the interactive web's content creation, collaboration and exchange by individuals that forming a virtual community on social media platforms (Zeng & Gerritsen, 2014).

Social media are fundamentally different as well compared to any traditional media because of their social network structure and egalitarian nature (Peters et al, 2013). Page 15 of 108

Peters et al (2013) argued that a brand is just a social actor in the network, which has no authority in hierarchical ‘1: n’ structure. In other words, a brand is just like other users in the network and no longer enjoy the traditional marketing privilege in imposing commercial messages exposure as in other media. Or merely a peer to peer communication channels which enable collaboration and exchange by its users (Zeng & Gerritsen, 2014). The following section will discuss the role of marketing in social media.

Hainla (2017) recently published that social media is continuing to be trending and growing. Taking Facebook as an example, in 2015, Facebook had successfully influenced 52% of the consumers on online and offline buying behaviour as compared to 36% back in 2014. In facts, there are 50 million small businesses worldwide are actively using social media channel such as Facebook to connect with their customers. A staggering of over 4.4 million videos were uploaded to Facebook directly back in February 2016, and successfully generated over 199 billion views from social media users worldwide. On different platform, Twitter has 328 million active users, and it is very active among its members in tweeting. The power of Twitter has proven that over 78% of people took the complains to a brand via Twitter are expected to be answered within an hour. Instagram, the picture, and video posting social media platform are popular among the businesses to engage with their followers which expected to rise to 70.7% by 2017 (Parker, 2016). According to Parker (2016), over 80% of the users are from outside of the United States. Thus, it is not just a social media platform for sharing videos or pictures, but a marketing channel for businesses to engage with their followers.

2.2.2 Social Media Marketing

Social media is a powerful component of an integrated marketing channel that its platforms included blogs, microblogs, email, and social networking sites provides many benefits for both marketers and consumers to achieve marketing communication and branding social goals (Kaur, 2016; Smith, 2014). Social media Page 16 of 108

has become an alternative communication tool supporting existing relationships and activities in a joyful way that can enrich the users’ experience, said Kim and Ko (2010).

In contrast to traditional marketing efforts, Weinberg (2009) commented that social media penetration would be continuing to grow and Internet has become increasingly social. Thus, this is a continuing trend where users are turning to social media for information, then traditional methods. Weinberg further commented that with the shift of traditional media to online social media, the consumer trusts on traditional media forms has declined and not as effective as it was before anymore. This is due to shifting into a customer-centric direction than listening to brands only (Hutter, Hautz, Dennhardt, & Füller, 2013).

In comparison with traditional marketing and the emerging of communications, consumers are taking the lead role in marketing being an active creator and influencer. This phenomenon shifted some powers of brands directly to consumers in social media platform according to Hutter et al (2013).

Social media marketing enabled brands or individuals to promote their websites, products or services through the social media platform and reach out to a larger user community. The form of promoting is through reviewing the content or finding a useful content and promoting it within the social media. The efforts emphasised collectively rather than individualised as in traditional marketing channels (Weinberg, 2009).

Hutter et al (2013) argued that the consequence of social media marketing campaigns on consumers’ perception of products and brands, and as well as the influence on purchase intention required more studies for better understanding. Despite the risks and doubts, brands cannot take the risks being absent in such powerful communication channel. Despite that, Zeng and colleague (2014) argued

Page 17 of 108

leveraging the social media to market products has proven to be an excellent strategy because of itself as an important tool. And thus, brands continuously investing in social media activities for being part of it and marketing their products and services. It is impactful as Tey (2015) wrote that about 70% of Malaysian turn to brands’ social media platform to research about the products before committing a purchase.

2.3 Brand Loyalty

To understand brand loyalty, it is important to derive brand from higher meta-level. According to Keller (2013), a brand can be rational and tangible, or more symbolic, emotional, and intangible. Keller (2013) further explained that to be tangible, it can be related to the performance of the product; while intangible more related to what the brand represents. Brands are important to both consumers and companies. For consumers, brands help to identify the maker of the product, take on special meaning for the consumers. Through some knowledge about a brand, consumers do not need to spend more time in researching and getting to know the brand. While for companies, a brand has a unique association and meanings that differentiate itself from other products. The quality and satisfaction of the brand will translate into brand loyalty that makes it more difficult for other companies to enter the market (Keller, 2013).

While brands matter, it is important to further explore the brand equity as an added value for companies in marketing concept (Vahdati & Mousavi Nejad, 2016). Aaker (1991) defined that brand equity is multidimensional theories and it includes attitudinal and behavioural dimensions, which consist of brand loyalty, brand awareness, brand associations, perceived quality, and brand proprietary. Brand equity has been studied in financial and consumers’ perspective. The previous perspective is focusing on the market value of the company’s asset, while the latter one reacts towards the brand name (Severi & Ling, 2013). Vahdati and Nejad (2016) Page 18 of 108

argued that brand loyalty played a significant role in brand equity. Hence, this section would further discuss on brand loyalty while other dimensions would be interests in future research.

Brand loyalty can be conceptualised as the final dimension of consumer brand resonance symbolising the with a brand, said Devi and Menezes (2016). Devi and Menezes (2016), and Severi and Ling (2013) commented that brand loyalty bring in sales, market share, profitability, helping business to the group or maintaining themselves in the marketplace is a result where brands have gain exclusive, positive, and prominent meaning in the minds of large numbers of consumers. The process of brand loyalty creation involved several variables amongst are the brand attachment and community commitment, where community members’ commitment to the community results in additional attachment to the brand around which the community is centred (Balakrishnan et al, 2014). Constant purchasing of one brand over time and declining to shift to competitors’ brands is a sign of brand loyalty (Severi & Ling, 2013; Calvo-Porral & Lang, 2015; Sasmita & Mohd Suki, 2015).

Branding has become more open and consumers are choosing which brands to interact with and shaping the brand values themselves (Smith, 2014). Smith further explained that consumers emotionally invest in the brand, more committed and loyal to the brand whom they are involved in the brand communities through the brand fan pages. Erdoğmuş & Çiçek (2012) commented that companies have better advantages when using social media marketing in social media platform to engage and build brand loyalty with consumers over the traditional platforms. In other words, social media helps brands to build brand loyalty. And it can be easily associated with the electronic word of mouth that peers review and conversation provide better credibility and thus boost the brand recognition and brand loyalty.

Page 19 of 108

2.4 Word of Mouth and Electronic Word of Mouth

Consumers are receiving high numbers of information regarding the brands, products, and services daily. The huge amount of information is through many advertisements advertised on different marketing channels. With many information at hands, it creates a difficult situation for consumers to decide on (Erkan, 2016). Therefore, WOM steps in as one of the most reliable sources of information for consumers since it consists of trusted consumers opinions and experiences than marketer advertised (Erkan, 2016; Huete-Alcocer, 2017; Gfrerer & Pokrywka, 2012; Nakayama, Wan, & Sutcliffe, 2010).

2.4.1 Word of Mouth

Word of mouth or WOM is considered one of the oldest methods of passing on information (Huete-Alcocer, 2017). It is defined in a form where verbal communication took place between person to person in a non-commercial context (Nakayama et al, 2010). Nakayama et al (2010) explained that verbal communication is inclusive of exchange of comments, thoughts, and ideas.

Gfrerer and Pokrywka (2012), and Huete-Alcocer (2017) presented that WOM is a widely accepted notion in the discipline of consumer behaviour that shaping consumers’ attitudes and behaviours toward the brands, products, or services. The communication between consumers is considered independent and non-commercial driven on the interpersonal exchange of information for the product or service. In other words, consumers always make the decision to buy on the advice given via word of mouth rather than trusting the advertiser’s promises. In other words, WOM is considered as a high level reliable, trustworthy, and credible source of information when compared to advertised media as a less reliable source. Products reviews and testimonies from known sources (the example of family members, relatives, peers, acquaintances, and friends) are considered trusted source in WOM Page 20 of 108

and is highly persuasive in altering consumers attitudes and behaviours. Besides that, the level of knowledge which the sender or the source of WOM plays an important role too. It happens when the receivers perceived that they acquired qualified information and convinced. Therefore, it will affect the consumers, or the receivers desired for purchases (Erkan & Evan, 2016). The influence of WOM is even stronger when consumers considering buying complex, durable, expensive, or critical products such as house, car, financial and healthcare services (Nakayama et al, 2010).

The reason consumers engaging in WOM is due to the feelings consumers’ experienced. Consumers that feels satisfied or dissatisfied are more inclined to engage in WOM activities. Consumers who experience positive feelings have the predisposition to share information supporting their brand. In other words, these consumers are considered as the loyal consumers (Gfrerer & Pokrywka, 2012). Hence, the sharing of experiences in WOM are considered genuine and unlike advertisements created by marketers that is not persuasive.

2.4.2 Electronic Word of Mouth

Electronic word of mouth is the transformation of traditional word of mouth as how people interact in traditional word of mouth (Hodza, Papadopoulou, & Pavlidou, 2012), but on a platform over the Internet (Erkan & Evan, 2016; Gfrerer & Pokrywka, 2012; Huete-Alcocer, 2017; Nakayama, et al, 2010). In contrast to WOM, eWOM is written form of the WOM on the Internet through the social media platform, and online user communities’ sites. eWOM functioned as a newer communication form to widespread the product reviews and testimonials to many consumers in a shorter time compared to WOM. It has the opportunity as well to spread the initial message online that considered unlimited reach. And it is considered reliable, credible, and trustworthy since the consumers have no commercial interests towards the product or the brand (Erkan & Evan, 2016; Gfrerer & Pokrywka, 2012; Huete-Alcocer, 2017; Nakayama et al, 2010; Hodza et al, 2012; Page 21 of 108

Smith, 2014; Balakrishnan et al, 2014). Although it is believed to be trustworthy and originated from consumers without self-interests, electronic word of mouth can be motivated and influenced by marketers to guide consumers’ in posting comments in the way they want it (Erkan & Evan, 2016; Hodza et al, 2012).

The internet is continuously evolving with the eWOM as communication become faster, easier with technologies. Consumers can search for marketer generated information and as well consumer-generated information for the product they intend to buy. Indeed, eWOM has the better advantage over traditional WOM because both consumers and marketers can post visual elements such as pictures, and videos that give other consumers an opportunity to better understand the information, and in turn helping them to make better decisions (Erkan & Evan, 2016). Additionally, eWOM provides companies the platform to identify the consumers’ needs and perceptions in cost-effective manners (Huete-Alcocer, 2017). Huete-Alcocer (2017) further argued that eWOM has become an important channel in social media marketing.

2.4.3 Differences between WOM & eWOM

One of the main differences between WOM and eWOM is the consideration of eWOM to be the electronic versions of traditional WOM. However, the comparisons suggested more than that. Huete-Alcocer (2017) argued the credibility of in circumstances involving many senders and receivers, and affecting the end consumers. The credibility of the sender’s message or information in WOM could be altered or distorted during passing the information to the next receiver and ultimately the end consumer. While in eWOM, the original message is in written form such as reviews, comments, or testimonies, reduced the risks as the original message can be traced back to the sender and increased the consumer confidences. However, credibility in WOM is highly positive as well when the receiver knows the sender in person which is considered as the trusted source of WOM (Gfrerer & Pokrywka, 2012; Huete-Alcocer, 2017; Nakayama et al, 2010). In contrast, sender Page 22 of 108

in eWOM could stay in anonymous, thus reduced the credibility of the relationship between both the sender and receiver. Huete-Alcocer (2017) further commented that privacy in WOM between the sender and receiver is high while in eWOM, information shared no longer is private and can be seen by anyone over the internet.

The speed and ease of access to information played a significant role to distinguish both WOM and eWOM. In eWOM, information can be obtained more quickly without the need to wait for the trusted source to offer an opinion. Besides that, it can be used to validate the received WOM reviews. As WOM communication occurs person to person, communication in eWOM is not bounded by time and location restrictions (Erkan & Evan, 2016; Huete-Alcocer, 2017).

Although both WOM and eWOM are seemingly similar, and at the same time very dissimilar. It is agreed that both are highly influential on consumer purchase behaviours.

2.5 Purchase Intention

Intention to purchase has been verified in many diverse research situations by researchers. As an example, Lim et al (2016) tested the purchase intention as mediating role in online shopping behaviour; while Danieli (2016) research on social media marketing system aiming to increase purchase intention. Besides that, Chang, Dewanti, and Pratiwi (2016) look into the brand and purchase intention; while Praharjo, Wilopo, and Kusumawati (2016) research on the impact of electronic word of mouth on repurchase intention mediating by brand loyalty and perceived risk.

Page 23 of 108

Purchase intention is expressed as consumer’s will to promise certain activity related to future consumption or possibility of purchasing in the future (Kim & Ko, 2010). Balakrishnan et al (2014) further explained that purchase intention carries several meanings which are the willingness to consider buying, an intention to buy in the future, and the decision to repurchase. Beyond that, purchase intentions refer to the degree of perceptual conviction of a customer to buy a product. Kim and Ko (2010) mentioned that purchase intention is depending on several factors such as cue that triggers consumer towards considering a product or a brand to be included in their consideration set. Consumers will undergo a process of recognising the intended product to purchase, then will further research and analyse the product before commit to buy it (Lim et al, 2016).

Erkan (2014) argued that information quality, information credibility, and information usefulness are the several factors influencing consumers’ purchase intentions. He concluded that higher quality reviews led to having stronger purchase intentions; while the positive relationship was observed between credible information towards purchase intentions. Furthermore, the usefulness of information is significantly associated with purchase intentions. Although the above-mentioned factors are influencing consumers’ purchase intention, Tariq, Abbas, Abrar, and Iqbal (2017) pointed out that purchase intention is increased with price and promotion, brand recognition, and as well the increased awareness with the brand.

According to Ajzen (1991) in the Theory of Reasoned Action (TRA), individual’s behavioural intention is affected by the perceived attitude toward the behaviour, and the subjective norm, the perception of social pressures from the people surrounding them. Hansen, Jensen, and Solgaard (2004) explained that TRA is concerned with rational volitional and systematic behaviour because most behaviours are social relevance and under volitional control and predictable from intention which the individual has control.

Page 24 of 108

Ajzen (1991) explained that the attitude toward the behaviour describes the degree to which a person has a positive or negative evaluation or appraisal of the behaviour in question. While, the subjective norm is a function of a set of beliefs termed normative beliefs that refers to the perceived social pressure to perform or not to perform the behaviour (Ajzen, 1991; Yousafzai, Foxall, & Pallister, 2010). As subjective norm describes the influence or the pressure of society on individuals. Electronic word of mouth is a way of displaying the potential, current or future comments either favourable or unfavourable regarding a product or brand online (Cheng & Huang, 2013) that lead to purchase intentions.

Further strengthen the Theory of Reasoned Action, Theory of Planned Behaviour (TPB) model posited better clarity that a person is assumed to perform certain behaviour if the person has actual control over the behaviour (Lim et al, 2016). Hence, when a person has favourable attitudes and subjective norm, it will lead to an acceleration of purchase intentions of a product or service.

2.6 Previous Empirical Studies

Empirical studies are the scientific method of testing a hypothesis through the collection and analysis of primary data that based on quantitative and qualitative approaches. The importance of empirical research is necessary to prove beyond the theory and abstract reasoning highlighted by researchers, academics, and scholars (Better Thesis, n.d.; Ithaca College Library, n.d.).

Therefore, in this section, the research will highlight previous empirical studies conducted by researchers to draw attention to the statistical associations between variables (Better Thesis, n.d.).

Page 25 of 108

2.6.1 Social Media Marketing to Brand Loyalty

Devi and Menezes (2016) conducted a structured questionnaire with 281 active social media users, and followed at least a brand on social media platforms. The results of the study show that social media marketing campaigns formed positive attitudes towards brand loyalty which is positively significant.

In 2015, Zheng, Cheung, Lee, and Liang (2015) conducted building brand loyalty through user engagement in online brand communities in social networking sites targeted 185 active Facebook users and collected the data through an online survey. The outcome of the study shows that user engagement influenced brand loyalty both directly and indirectly.

Erdoğmuş and Çiçek (2012) studied the impact of social media marketing on brand loyalty in Turkey on customers who followed at least one brand on the social media. Questionnaire was used to collect data from 338 people. The results of the study shown that customers’ brand loyalty was positively affected by the social media marketing offered by the brand.

2.6.2 Electronic Word of Mouth to Brand Loyalty

Severi et al (2014) conducted the effect of electronic word of mouth on brand equity back in 2014. The study collected a total of 300 useful data via questionnaires in one of the private university in Malaysia. The outcome of the study illustrated that electronic word of mouth significantly impacted the brand loyalty in this specific association.

Balakrishnan et al (2014) studied the impact of social media marketing medium towards purchase intention and brand loyalty among generation Y in Malaysia. The Page 26 of 108

test results were significant where electronic word of mouth has a positive impact on brand loyalty. This study was conducted via the questionnaire with 200 students.

2.6.3 Social Media Marketing to Purchase Intention

Danieli in 2016 applied mixed method design to gather and analyse both qualitative (email interview) and quantitative data (online survey) on the social media marketing strategies towards the increase of consumers’ purchase intentions. Through her finding, social media marketing is found to be influential upon purchase intentions.

Mao, Sang, and Zhu (2014) studied the consumer purchase intention based on social media marketing with a total 283 questionnaires taken back. From the analysis, social media marketing positively influences the consumers’ purchase intention

2.6.4 Brand Loyalty to Purchase Intention

In 2015, Calvo-Porral and Lang (2015) conducted an online survey in Spain for private labels which involved the association between brand loyalty and purchase intention. A total of 362 valid response were used to analyse and there is a high influence of brand loyalty towards consumers’ purchase intention.

Chang et al (2016) studied the success of Toto brand in grabbing purchase intention in Indonesia. The researchers conducted 200 surveys and found that brand loyalty has a significant impact on consumers’ purchase intention.

Page 27 of 108

2.6.5 Electronic Word of Mouth to Purchase Intention

Tariq et al (2017) analysed the impact of electronic word of mouth and brand awareness on consumers’ purchase intention in Pakistan. The questionnaire was used and collected 262 responses of mobile users. The correlation between the electronic word of mouth and purchase intentions is significant and positively correlate with each other’s.

Erkan and Evans (2016) conducted a survey of 384 university students who are using social media website to investigate the influence of electronic word of mouth in social media on consumers’ purchase intention. The outcome of the study shows that electronic word of mouth positively affects consumers’ purchase intentions.

Jalilvand and Samiei (2012) from Iran conducted a study in 2012 through the survey with 341 respondents who had experience within online communities. The study of the impact of electronic word of mouth on brand image and purchase intention in the automobile industry proven that electronic word of mouth is one of the most effective factors in influencing consumers’ purchase intentions.

2.7 Theoretical Framework

2.7.1 Theory of Reasoned Action

The theory of reasoned action (TRA) is a well-known social psychological model developed by Fishbein and Ajzen (1975). This model assumed individuals are rational and systematically make use of information available for them. The

Page 28 of 108

individual will consider the implications of their actions before deciding to engage or not engage in each behaviour (Hansen et al, 2004; Yousafzai et al, 2010).

Figure 1 Theory of Reasoned Action

Figure adopted from Octav-Ionut (2015)

According to TRA model (Figure 1), an individual’s behavioural intention is a function of two determinants: a personal factor termed attitude toward the behaviour, and a person’s perception of social pressures termed subjective norm, according to Hansen et al (2004) and Yousafzai et al (2010). In other words, the behaviour intent is deemed the most important determinant of a person’s behaviour.

The individual’s attitude towards behaviour includes evaluation of behavioural outcomes, and behavioural beliefs. According to Li (2010), attitude is an individual positive or negative emotional state about performing the target behaviour. In other words, if the results of performing a behaviour are positive, the individual will have positive attitude performing that behaviour. Likewise, if an individual believes that the behaviour would lead to an undesirable result, the individual would have a negative attitude towards it (Lezin, n.d.).

Page 29 of 108

While the subjective norm consists of normative beliefs, and motivation to comply. Lezin argued that subjective norms are perceptions of the beliefs influenced by people surrounding the individual. Subjective norm is a function of a set of beliefs termed normative beliefs that refer to the perceived social pressure to perform or not to perform the behaviour (Ajzen, 1991; Yousafzai et al, 2010). Thus, the individual would question if the individuals or groups would approve the behaviour or not. And it also depending on whether the individual would comply with the other views or opinions. As subjective norm describes the influence of society on individuals, eWOM is a way of displaying the subjective norm which it describes the potential, current or future comments either favourable or unfavourable regarding a product or brand online (Cheng & Huang, 2013).

The TRA is a general model and, as such, it does not specify the beliefs that are operative for a behaviour (Yousafzai et al, 2010). TRA deal with prediction rather than the outcome of behaviours as researchers are often interested in situations in which the target behaviour is not completely under the individual’s control Yousafzai et al, 2010; Hansen et al, 2004). Another word, TRA is uncappable to predict the situation in which individuals have a low level of volitional control.

2.7.2 Theory of Planned Behaviour

According to Li (2010), Theory of Reasoned Action was applied to predict individual behaviour in a voluntary situation, not in a compulsory situation. To tackle the limitation of the TRA model, the theory of planned behaviour (TPB) was introduced by Ajzen to deal with behaviours which people have incomplete volitional control (Ajzen, 1991). In the extension of the theory, perceived behavioural control was added as the third construct that influences behavioural intentions and actual behaviours (Yousafzai et al, 2010) as in Figure 2.

Yousafzai et al (2010) explained that the consideration of control beliefs is about the resources and opportunities possessed by the individual and the anticipated Page 30 of 108

impediments towards performing the target behaviour. Cheng and Huang (2013), Lim et al (2016) explained that when a person perceives that they have the ability, resources, and opportunities to perform a certain behaviour, they will perceive themselves as more in control and their behavioural intention will increase too.

However, Madden, Ellen, and Ajzen (1992) argued that perceived behavioural control is motivated towards behaviour through behavioural intention. The assumption believes that if the individual has little control over performing the behaviour because lack of requisite resources, then the intentions to perform the behaviour may be low even if the individual has favourable attitudes and/or subjective norms.

Ajzen (1991) summarised that attitude and subjective norm which are more favourable and with respect to a behaviour, there are two effects of perceived behavioural control. First, perceived behavioural control is motivated to have an indirect effect on behaviour through intentions. Second, perceived behavioural control reflected the actual control and has a direct connection to the behaviour that was not mediated by behavioural intentions (Madden et al, 1992).

Page 31 of 108

Figure 2 Theory of Planned Behaviour

Figure adopted from ASU (n.d.)

2.8 Conceptual Framework

In this study, the Theory of Planned Behaviour (TPB) proposed by Icek Ajzen in 1985 links beliefs and behaviour together to improve on the predictive power of the theory of reasoned action including perceived behavioural control. The conceptual framework seeks to find out the relationship between the independent variables and as well the effect on the dependent variable which is the purchase intention.

Social media as the new frontier is the platform of two-way communication that allows users to interact with each other online and share information and opinion. It is not going away anytime soon, and businesses must get on to leverage the power of social media or left behind. The Millennials are the technology savvy cohort that seeking out the best brand options for themselves amid countless choices than sticking to one brand only. Consumers have turned into social media to share their personal experiences towards a product, service, and brand. And it results in Page 32 of 108

affecting others positively or negatively towards the product, service, or the brand after information posted on social media sites. Businesses increased brand related interactions and marketing campaigns exposure through social media to gain exclusive, positive, and increasing transparency towards the brand, product, or services.

Brand loyalty is one of the attributes in brand equity describing the pattern of consumer behaviour where consumers become committed to brands and lead to purchase from the brand. Brand loyalty can bring in sales, profitability, increase or maintain market shares. While brand loyalty consistently linked to repeat purchases from the same brand over time. However, brand loyalty creation is not necessary a post-purchase behaviour but previous experiences towards the brand that leads to brand attachment.

Electronic word of mouth (eWOM) is using the internet as a platform compared to traditional word of mouth through the statement either positive or negative made by potential, actual or former customers about a product or company, which is made available to a multitude of people and institutions (See-To & Ho, 2014).

Page 33 of 108

Figure 3 Conceptual Framework

2.9 Hypotheses Developments

In this section, there are five hypotheses to be presented based on each relationship established developed through conceptual framework discussed in previous section.

2.9.1 Social Media Marketing and Brand Loyalty

Devi and Menezes (2016) pointed out in their study, the beneficiary campaigns on social media are the most significant drivers of brand loyalty. Growing power in consumers instead of brand marketers and deep understanding of consumer perceptions of social media activities is crucial for engagement and trust of brand (Smith, 2014). Kim and Ko (2010) further commented that understanding consumers’ purchasing behaviour is important to a business to attract and retain its customers.

Page 34 of 108

Consumers perceived that social media marketing is sincerer as it is trying to show what the brand is about rather than controlling its image (Erdoğmuş & Çiçek, 2012). Erdoğmuş and Çiçek (2012) elaborated that networking, conversation, and community are amongst the factors helping firms to build brand loyalty via social media with small campaigns that making connections and relationship with the consumers in a very short time. The popularity of the social media platform and the contents are important reasons consumers engaging with brand on social media beyond the traditional methods as explained by Erdoğmuş and Çiçek (2012). Effective campaigns on social media took the consideration of the benefits, values, and advantages firms had offered to the consumers that will incrementally increase the effect on brand loyalty through satisfaction (Erdoğmuş & Çiçek, 2012; Balakrishnan et al, 2014). Balakrishnan et al (2014) stated that the consumers’ commitments and engagement to the social media results in additional attachment to the brand which leads to positive word of mouth and purchase intention.

H1: Social media marketing will have significant positive influence on consumers’ brand loyalty.

2.9.2 Electronic Word of Mouth and Brand Loyalty

eWOM is like traditional word of mouth as how people interact (Hodza et al, 2012) but eWOM is faster and far-reaching to multitude of people and it is perceived trustworthy because it is originated from consumers who have no personal interests in the product or the company (Balakrishnan et al, 2014; Hodza et al, 2012; Smith, 2014). Laroche, Habibi and Richard (2013) mentioned that enhancing the relationships in the brand community should increase the brand trust which has a positive effect on brand loyalty.

Page 35 of 108

Social media marketers using social media to communicate information about products, while consumers post their product experience or review and as well questions related to the product (Smith, 2014). The relationship enhanced when sharing and dissemination of information will reduce the information asymmetry, uncertainty and increase the brand’s predictability which resulting in trust enhancement (Laroche et al, 2013). Balakrishnan et al (2014) stated that eWOM can provide the opportunity to consumers to read about others opinion and experiences as well as the contribution by themselves, it has positive impact on brand satisfaction that will lead to brand loyalty. Laroche et al (2013) argued that their research was consistent with other studies where participation in social virtual communities positively influences the brand loyalty. Likewise, Kim, Sung and Kang (2014) implied that the consumers’ intention to engaging with the brand community to certain extent represented the satisfaction towards the brand.

H2: Electronic word of mouth will have significant positive influence on consumers’ brand loyalty.

2.9.3 Social Media Marketing and Purchase Intention

Social media as described in the earlier literature is a powerful component of integrated marketing channel and alternative communication tools supporting the existing relationship between marketers and consumers with interactions and enrichments (Kaur, 2016; Kim & Ko, 2010; Smith, 2014). The encouragement of social interactions among consumers on social media such as communities, reviews or recommendations is likely to establish trust among themselves (Hajli, 2014). Hajli (2014) stated that consumers’ socialisation through its communities on social media is, directly and indirectly, generating social support that influencing trust towards a product or brand. The exchange of information, feedback, and ratings on social media platform overcome the barrier of insecurity and increase the trust of the consumers and consequently lead to the willingness to purchase the product or service (Hajli, 2014). Page 36 of 108

H3: Social media marketing will have significant positive influence on consumers’ purchase intention.

2.9.4 Brand Loyalty and Purchase Intention

Smith (2014) stated that brand fan pages facilitated the relationship between consumer and brand that leads to stronger purchase intention while Severi and Ling (2013) stated that brand loyalty is important when evaluating a brand in values because loyalty can generate profits. The meaning of cognitive loyalty would describe when the consumers’ need to make a purchase decision, the first brand that comes up in mind is the consumer first choice (Balakrishnan et al, 2014). Likewise, it is applicable to purchase intention where the first brand comes up in the consumer mind when thinking of a product or service. Because brand loyalty has the power to influence consumer decision to purchase the product, service or brand and decline to shift to other brands. Therefore, brand loyalty has the intention to influence consumers in purchase intention.

H4: Brand loyalty will have significant positive influence on consumers’ purchase intention.

2.9.5 Electronic Word of Mouth and Purchase Intention

Kaur (2016) argued that over the last few years, users have spent more time and shared more information, thoughts, and opinions with each other easily over the Internet. eWOM played a significant role not that, it has an impact on the cocreation value, and because of that it has an impact upon purchasing intention regardless the type of virtual communities’ consumers is in (Danieli, 2016; Balakrishnan et al, 2014). eWOM generates a greater effect on purchase intentions compared to advertisements because members of the communities are sharing their knowledge or experiences that others viewed as unbiased and trustworthy Page 37 of 108

information (Balakrishnan et al, 2014; Hodza et al, 2012; Smith, 2014). Therefore, the hypothesis suggested eWOM will have the significant influence on consumers’ purchase intention.

H5: Electronic word of mouth will have significant positive influence on consumers’ purchase intention.

2.10 Summary

In this chapter, each of the variables for this research was reviewed to get a comprehensive understanding of the secondary information. The social media marketing, brand loyalty, and electronic word of mouth are the main independent variables that are discussed together with purchase intention as the dependent variable. Additional information such previous empirical studies did by researchers were shared to prove beyond the theories and abstract reasonings. Through the discussion of the theoretical framework, the conceptual framework was created. The conceptual framework was designed in according to the research objectives and questions direction in Chapter 1 in order hypotheses can be developed accordingly.

Page 38 of 108

CHAPTER 3

METHODOLOGY

3.1 Introduction

This chapter outline the methodology used in this research to collect the data for later analysis through statistical software SPSS. Thus, this chapter will explain the research designs, types of research methods, sampling criteria, the instruments used to collect data for later analysis in statistical software and follow by type of statistical testing to provide a better understanding for later analysis in the next chapter.

Page 39 of 108

3.2 Research Design

A research is defined as a scientific process to validates and refines existing knowledge and generates new knowledge (Cham, n.d.a). A research design is a blue print or the framework that has been created to answers to the research questions. Sahay (2016) emphasised that the planning and designing were the most important in a research through Saunders’s research ‘onion’. Whereby, the outer layer forms the root while middle layers are the building block of the research. The importance also shared by Wilson (2014) that it helps to clarify the research designs, to recognise which designs work best, and assist the researcher to identify and adapt research designs accordingly.

Saunders, Lewis, and Thornhill (2012) through the research ‘onion’ claimed that researchers need to address several aspects such as research philosophy, approaches, methodological choice, research strategies, research methods, time horizon, and the research techniques and procedures. The benefits of the research ‘onion’ create a series of stages and the steps by which a methodological study can be described. Which is coherent with the objectives and research questions of the research (Sahay, 2016).

Similarly, Wilson (2014) introduced the Honeycomb of Research Methodology which shares the similarity to Saunders et al (2012) research ‘onion’. Wilson’s research methodology counter other research models that do not always indicate the link between each element. The Honeycomb of Research Methodology as well recognised the fact that the thought process may not necessarily be linear all the times.

There are two major ways of thinking in the research paradigm namely ontology, and epistemology that create a holistic understanding of how knowledge is regarded, and in what manner we are capable of seeing ourselves in relation to this knowledge Page 40 of 108

(Patel, 2015; Saunders et al, 2012). Ontology is assumed relates to the nature of reality and its characteristics (Boateng, n.d.; Saunders et al, 2012). In another way, how the researcher perceives the world is (Patel, 2015). According to Saunders and colleagues (2012), objectivism and subjectivism are the two aspects of ontology. Whereby, objectivism portrays the position of the social entities exists in reality and being independent external to those social actors. In another word, objectivism implied that social phenomena are based on independent realities that are beyond the grasp or control.

On contrary, subjectivism asserts that social phenomena are a result created through the perceptions and conducts of the social actors (Saunders et al, 2012). Hence, researchers conduct their study with the intention of understanding what is happening and reporting these multiple realities (Boateng, n.d.; Creswell, 2007; Saunders et al, 2012). Wilson further commented that subjectivism is linked to interpretivism where researcher examines the motivation and social interactions of the respondents.

Epistemology, on the other hand, construed what is considered the acceptable knowledge in the field of study (Saunders et al, 2012; Wilson, 2014). Concisely, the relationship between the researcher and what is known. In the positivist approach to the study, the researcher is independent of the research. Which means that researcher is objective and is detached from those involved in the study (Wilson, 2014). The research strategy would involve testing and confirm the developed hypotheses. According to Saunders and colleagues (2012), only phenomena that positivist can observe, or measure will produce credible data (Boateng, n.d.). That’s mean, in carrying out the research, the approach used would be a deductive approach. Whereby, observations are quantifiable compared to qualitative and the finding is applicable to the whole population (Wilson, 2014).

In contrast, interpretivist has different views from positivist. Unlike positivist which generalise their subject, interpretivist goes in-depth into their research subject with Page 41 of 108

high levels of interaction and participation (Wilson, 2014). According to Saunders, Lewis, and Thornhill (2012), interpretivist needs to understand the differences between the subjects, and continuously interpreting the social world around them that lead to adjustment of their own meaning and actions. Similarly, Boateng (n.d.) commented that knowledge is established through the meanings attached to the studied phenomena. In another word, the researcher is not interdependent with the research subjects. Thus, the research is based on inductive approached that.

Zikmund, Babin, Carr, and Griffin (2010) explained that besides classifying research on the basis of technique, classifying the research by purpose show how the nature of a decision would influence the research methodology. They further argued that matching a certain decision circumstances with the right form of research is imperative in getting valuable outcomes. There are three types of research classified as exploratory, descriptive, and causal.

Exploratory research is the first step of a research where often used as a guide and further refine the requirements for subsequent research efforts (Zikmund et al, 2010). Sekaran (2003) argued similarly that exploratory research is required when the researcher has no information how the research issues have been solved in the past or limited known facts available, therefore needed more information to develop a viable theoretical framework. Typically, the exploratory research is an unstructured approach with the flexibility to adapt to changes as needed (Saunders et al, 2012). Saunders and colleagues (2012) further argued that researcher undertaking exploratory research must be willing to change the direction as of when new data and insights made available.

Descriptive research is often undertaken to ascertain and able to describe the characteristics of the research variables in a situation (Saunders et al, 2012; Sekaran, 2003; Zikmund et al, 2010). The descriptive research is contrary to exploratory research, where researcher conducted the research after gaining a firm understanding of the studied situation. Sekaran (2003) listed that descriptive Page 42 of 108

research helps to understand the characteristics of the variables in the given situation, to have a systematic mindset about aspects in given situation. Further to that, Sekaran (2003) also commented descriptive research offer ideas for further probing and research. And it also helps to make certain simple decisions.

The causal research involved hypotheses testing that required to explain the nature of certain relationships. In other words, causal research wants to identify the cause and effect relationship whereby the effect is the outcome of the research (Zikmund et al, 2010). It is also undertaken to explain the variance in the dependent variable or to predict organisational outcomes (Saunders et al, 2012; Sekaran, 2003). Zikmund and colleagues (2010) expressed that exploratory research is the foundation for descriptive research while it establishes the basis for causal research.

In this research, the research will adopt the positivist stance where it will apply the objectivism from the ontology perspective. This also aligned with the epistemology perspective that the research is done through structured data collection via questionnaire and analyse via statistical software.

3.3 Research Approach

After the understanding of the philosophies or theories, the research raises the important question on the research approach. According to the research ‘onion’, this layer will further discuss the reasoning of deductive and inductive.

Deductive reasoning occurs when it commences with and applies a known theory rather than endeavouring to create a new theory (Boateng, n.d.; Saunders et al, 2012; Wilson, 2014). In another word, deduction comprises the development of theory into hypotheses that go through series of rigorous testing. Therefore, there is a Page 43 of 108

causal relationship between concepts and variables that researchers keen to seek for explanations. Wilson (2014) elaborates where theory and hypotheses come first in the research and will continue to influence the rest of the research process. Most of the time, the deductive approach in research is often related with the quantitative type of research. Hence, there is a necessity to select a sufficient sample size to take a broad view of the deductions through statistical analysis and a confirmation of the original theories (Trochim, 2006a). Therefore, it is a highly structured approach (Gill & Johnson, 2010; Saunders et al, 2012; Wilson, 2014).

On the opposite spectrum, the inductive approach works differently from the deductive approach. Inductive reasoning is undergoing a theory building process that begins with observations and to establish a theory through the generality of the phenomenon (Wilson, 2014). According to Saunders and colleagues (2012), inductive approach is particularly concerned with the context in which such events were taken place where only small sample size will be studied. Therefore, it is often associated with qualitative research. This would allow flexible structure approach to permitting changes of research emphasis as the research progresses (Wilson, 2014).

From a different perspective, Sekaran (2003) described similarly to the research approach as hypotheses testing and descriptive study. Sekaran (2003) commented that hypotheses testing are usually testing the relationships or establish the differences among the factors in a circumstance that explain the variance in dependent variable or anticipating the organisational outcomes. While descriptive study is to ascertain and can describe the characteristics of the variable in the situation. The descriptive study allows the researcher to understand the characteristics of the subject in each situation, to further probe and research.

Page 44 of 108

3.4 Research Methodological Choices

Zikmund et al (2010) explained that the right approach needs to match to the right research context. And Zikmund and colleagues (2010) further commented that qualitative research can achieve research objectives that quantitative research cannot and vice versa. Saunders et al (2012), and Wilson (2014) agreed that qualitative and quantitative methods do not necessarily have to be exclusively used in research as many business and management research designs are likely merging both qualitative and quantitative methods.

Quantitative research often associated to positivism. However, Saunders and colleagues (2012) argued that quantitative research methodology can be used with other paradigms such as realism and pragmatism. The quantitative approach would draw a large sample from the population to measure the subjects that involved numerical measurement and construct sweeping statement regarding the population as an entire through statistical analysis (Zikmund et al, 2010; Wilson, 2014). Unlike qualitative research, quantitative approach is often associated with the deductive approach but may also incorporate with inductive approach when developing the theory (Saunders et al, 2012; Wilson, 2014).

As mentioned before, the characteristic of quantitative research is a highly structured approach where analysis is the statistical and involved analysis of the results following the theoretical applications (Saunders et al, 2012; Wilson, 2014). Saunders et al (2012) elaborated that in quantitative research, survey research strategy is widely applying the usage of questionnaires or structured interviews. Due to the application of theory which has been used by previous studies, the results of the study are comparable with the current studies (Wilson, 2014).

According to Zikmund et al (2010), qualitative research has interest in the qualities rather than quantities. In another word, a research that addressed research aims Page 45 of 108

through discovering true inner meaning and new acumens without reliant on mathematical measurements. As commented by Saunders et al (2012), qualitative research is highly correlated with interpretivism philosophy and it is used within realism and pragmatism paradigms.

As established in this chapter, qualitative research is widely associated with the inductive study (Creswell, 2007; Saunders et al, 2012; Wilson, 2014). However, quantitative research is used as well on deductive approach when testing existing theories (Saunders et al, 2012). Data collection often is not structured and practiced free-form. The qualitative researchers tend to get closers to the subjects’ perspective through a thorough interview and observation to establish patterns or themes (Creswell, 2007; Saunders et al, 2012; Wilson, 2014; Zikmund et al, 2010). Through the interpretation of the observations and interviews, new theories are proposed that illustrate the findings.

In research design, it is often seen that researchers applied multiple methods to answer the research objectives. In research philosophy of realism, realist may have applied quantitative analysis for data analysis and followed by qualitative research to explore perceptions. Similarly, for pragmatism, pragmatist valued both quantitative and qualitative research and the exact choice will be reliant on the nature of the research (Saunders et al, 2012). According to Wilson (2014), it is arguable that in the simplistic view, combining the quantitative and qualitative research are so-called mixed methods as when research involved different paradigms. However, Wilson (2014) further argued that quantitative and qualitative research are not combined but work in multiple levels fitting into the context as multi-strategy. In other words, a researcher would use methods that fit best for the study rather than fitting into any paradigm.

Saunders and colleagues (2012) branch out multiple methods into both multimethod and mixed methods. In multimethod research, the research analysis is independent and do not mixed between quantitative and qualitative research. A research that Page 46 of 108

collects quantitative data using questionnaires and structured observations for an example, the researcher would only use the quantitative statistical process to analyse the data. Likewise, for qualitative data collection, the data would be studied with qualitative measures. As opposed to multimethod, mixed methods combined both quantitative and qualitative research in the research design. A study would have combined variety of ways from simple, convergent forms to complex, fully integrated forms that involved a number of dimensions of mixed methods research.

While in this study, the approached applied is quantitative research which related to positivism philosophy and applying the deductive approach. Therefore, this study is considered structural approach whereby collected data will be analysed statistically with statistical analysis software. And to find answers in answering the research aims based on the theoretical application.

3.5 Population and Sampling

A population is referred to the entire group that not limited to people, events, or things that share some common set of characteristics the researcher wishes to study (Greener, 2008; Sekaran, 2003; Zikmund et al, 2010). In other words, it referred as the complete set of data which a sample is taken (Saunders et al, 2012). The idea of studying every element in the population is possible and known as census (Greener, 2008; Saunders et al, 2012; Zikmund et al, 2010). However, to draw some conclusion out from the population is impractical due to high cost, time constraint, and sampling data are sufficient for decision making (Cham, n.d.b). In the interest of the research, the target population is defined and identified at the beginning of the study. In this research, the sampling frame is not available as the target population is targeted any residents staying and living in Malaysia where it is unlikely available to obtain the list of resident’s names due to confidentiality.

Page 47 of 108

The sample is referred as a subset of the target population where the data collected from this subset is possible to generate findings to generalise the entire population (Greener, 2008; Saunders, Lewis, and Thornhill, 2012; Sekaran, 2003; Zikmund et al, 2010). As mentioned before, census study is extremely high cost and time constraint, sampling on the other way is faster and cheaper to conduct the survey. The accuracy of sampling lies on the factors of sample size such as bigger sample size to reduce error and the quality of the survey’s methods (Saunders et al, 2012). Zikmund and colleagues (2010) argued that sample on occasion can be more accurate than a census due to the interviewer mistakes on such high workloads.

Sampling strategy is categorised into two probability sampling techniques and nonprobability sampling techniques. According to Zikmund et al (2010), in probability sampling technique, every subject of the population is a known and has an equal probability of being selected. Sekaran (2003) further commented that the representative of the sample is important in the interests of wider generalisability. On nonprobability sampling, the probability of each subject being selected is unknown (Saunders et al, 2012; Sekaran, 2003), and rely on the researchers’ personal judgement (Zikmund et al, 2010).

In probability sampling, there are four techniques can be used to obtain a representative sample. Simple random, systematic sampling, stratified sampling, and cluster sampling are the techniques in probability sampling. These techniques are useful when sampling frame is available (Saunders et al, 2012). In nonprobability sampling, similarly, there are four techniques to obtain a sample. There is the convenience sampling, judgmental sampling, quota sampling and snowball sampling.

The selection for this study is on nonprobability sampling technique. The applied sampling method is convenience sampling. Convenience sampling is widely used, and it is as well easily to obtain compared to other methods (Saunders et al, 2012). Zikmund and colleagues (2010) echoed that convenience sampling is effective in Page 48 of 108

obtaining a large number of data quickly and economically. On the target population, this study will be focusing on respondents from Malaysia where subjects have access to the social media platform.

3.6 Sampling Size

In any research study, the consideration of study the entire population is an exhaustive of time, money, and manpower. Hence, sampling method would consider only a subset of the entire population available. Saunders and colleagues (2012) discussed that the larger the sample size, the lower the error when generalising the population. However, the sample size is dictated by the confidence level, margin of error that can be tolerated in the study, type of statistical analyses, and the total population size. In the majority of social science studies, researchers are content to estimate the population’s characteristics at 95% confidence to between 3 ~ 5% of margin error.

Saunders et al argued that useful rule of thumb of minimum 30 sample sizes for each category within the sample. While Hair, Black, Babin, and Anderson (2010) suggested a ratio of 5:1, which consist of 5 observations are made for each independent variable. However, Hair et al (2010) commented that the desirable level is between 15 to 20 observations for each independent variable.

Hence, in this study, based on Saunders, Lewis, and Thornhill (2012) requirement, a minimum of 150 sample sizes are needed for the 3 independent variables. While following Hair et al (2010) rules of thumb, a minimum of 100 sample sizes when considering 5 or fewer constructs.

Page 49 of 108

3.7 Data Collection Methods / Instruments

There is two type of data uses in research which known as primary and secondary data (Saunders et al, 2012). In order to answer the research questions for this study, data are collected through the primary source which is the personally administered questionnaire. There are several factors in favour of primary data over the secondary data in this study. Saunders and colleagues (2012) explained that the collected data may not match the purpose of this study and deem inappropriate to the research questions. The access to the data may be difficult due to data is available for commercial reasons and expensive. Although there are many secondary data available from reliable sources, however, there is not any real control over the data quality and how the initial purposes manipulate data.

The questionnaire is widely used to collect data from the survey strategy (Saunders et al, 2012). Saunders (2012) further commented that everyone is given to the same set of questions and provides an efficient way to collect response from a large sample for quantitative analysis. According to Sekaran (2003), the questionnaire is a preformulated set of questions to which respondents’ answers within closely defined alternatives.

In this study, personally administered questionnaire method is applied. One of the benefits where there is an opportunity to introduce the research topic and motivates the respondents to offer their frank answers (Sekaran, 2003). At the same time, it is less time consuming, less expensive, and does not require any skills to collect data. The structured questionnaire is designed and created by the Google Forms online platform (Google, n.d.). Google Forms is an internet-mediated or known as webbased questionnaires that are widely accessible through internet connectivity on various devices such as on computers or tablets/smartphone.

Page 50 of 108

The structured of the questionnaire is thoughtful design with the following layout. Section 1: Brief introduction the purpose to conduct the survey is for research purposes. Section 2: The questions were focused on demographic, social media, and brand consciousness which will lead the respondent to the next section. In Section 3, questions were focusing on the construct measurements for social media marketing, brand loyalty, electronic word of mouth, and purchase intention. A total of 25 questions were created for this study.

According to Saunders, Lewis, and Thornhill (2012), choosing the time horizon is important when conducting a research. There are two types of time horizon which are cross-sectional and longitudinal studies. Sekaran (2003) and Saunders et al (2012) argued that taking a snapshot is to study the phenomenon over the constraint time in order to answer the research question. Contrary, longitudinal studies involved a longer period and with the capacity to study change and development (Saunders et al, 2012). According to Sekaran (2003), data is collected at more than one point in time, therefore the study is not cross-sectional anymore. In this study, the approach applied cross-sectional studies for a period to complete the entire research due to limited time availability.

3.7.1 Pilot Study

Prior to the actual survey, a pilot test was conducted to validate the questions. The questionnaires were designed and asked in the simplest way possible to prevent confusions. The questions were reviewed by the expert in this area and went through several rounds of adjustment to seek better clarity and avoiding misinterpretation of the questions. Due to the questionnaires were not targeting on specific brand or product, a context was established to guide respondents towards the objectives of the survey.

Page 51 of 108

After the first round of validation, the questions were distributed to a group of friends from various professional backgrounds. The purpose of doing so was to ensure respondents able to understand the questions and provide the perceived answers in higher accuracy. Confusing terms and questions were identified and improved for better understanding. Furthermore, new questions were introduced as a guide to guide respondents to the next sections. The improved questions were tested again with a different group of friends that has no direct or indirect knowledge of the questions prior to the validation. The participants did not face any problem when reading the questions. The questions were confirmed and created in Google Forms for general survey purposes.

3.7.2 Construct Measurement

The questionnaire items were adapted based on numerous sources through literature reviews. The following table comprised the questions where it was measured on a 5-point Likert scales ranging from strongly disagree to strongly agree.

Table 1 Construct measurements Category Questions SM1

Source

Which of the following social media site do you have an account with? (Check all that apply)

SM2

How often do you use social media?

Savage, 2012

SM3

Are you brand conscious?

Savage, 2012

Example of brands: Adidas, 100+, Starbucks, Uniqlo, Petronas, AirAsia, Honda, etc.

Page 52 of 108

SMM1

There

are

many

marketing

campaigns Erdoğmuş

&

(advertisements, videos, images, posts, reviews, Çiçek, 2012 etc.) by the brand on social media site. SMM2

The brand regularly updates its contents (posts, Danieli, 2016 pictures, videos, etc.).

SMM3

The contents (posts, pictures, videos, reviews, etc.) Erdoğmuş are relevant to me.

SMM4

Çiçek, 2012

The contents (posts, pictures, videos, reviews, etc.) Erdoğmuş are popular among friends or others.

SMM5

&

&

Çiçek, 2012

The brand uses applications (mobile apps) and Erdoğmuş

&

different platform (social media, website, email, Çiçek, 2012 SMS, telephone, etc.) in promoting their products and services. BL1

When I need to buy a product/service, my first Yuan, 2015 thought is this brand.

BL2

I feel secure when I buy this brand because I know Shieh & Lai, that this brand will never let me down.

BL3

2017

I am willing to spend more time and to pay more if Saritas I am satisfied with this brand.

BL4

&

Penez, 2017

I feel loyal to the brand because they regularly Marketest, n.d. offer rewards (discounts, free gifts, etc.) to engage with me.

BL5

I have positive feelings about my favourite brand Jeng, 2016 and therefore plan to remain a consumer of my favourite brand.

EWM1

I often read online reviews (number of likes, Erkan shares,

comments,

ratings,

etc.)

on Evans, 2016

products/services from different brands, and I think they are generally informative.

Page 53 of 108

&

EWM2

I think the online reviews (number of likes, shares, Erkan

&

comments, ratings, etc.) are helpful for my Evans, 2016 decision making when I buy a product/service. EWM3

Recommendations by friend make me feel Adroit Digital, confident in buying the product/service.

EWM4

I will recommend the product/service to my Erkan friends or other consumers via social media.

PI1

Evans, 2016

Goh, 2017

I will buy the product/service next time when I Hsu, Chang, & need it.

PI3

&

The probability that I will buy the product/service Tan, Ooi, & is very high.

PI2

n.d.

Chuang, 2015

It is likely that I will continue buying a Hsu, Chang, & product/service from the brand in the future.

Chuang, 2015

3.7.3 Data Scale of Measurement

In this study, there are several measurement methods were applied in the questionnaire to collect gather the respondents’ information. In the construct of measurement sections, scaling is applied, and to be specific which is the Likert scale. According to McLeod (2008), the Likert scale is widely used to measure individual attitudes. Thurstone (1928) argued that distribution of attitudes of the sample on a specific topic may be represented in the form of the frequency distribution. Which means the entire range of opinions from one end who are most strongly in favour of the topic to the other end of the scale who are strongly against it. Likert scale was developed on the principle to collect respondents input to a series of statements on a topic in the extent to which they agree with them (McLeod, 2008). McLeod argued that the quantitative data collected can be analysed with relative ease compared to simple yes/no answers.

Page 54 of 108

While on the other part of the questionnaire, the focus on demography information applied nominal, ordinal, and ratio scales to collect respondents’ demographic information such as age, gender, education level, and employment status.

3.8 Variables

A variable can be explained as an entity which has the characteristics to change and can take on different values (Mustafa, 2015; Trochim, 2006b). An independent variable is defined as the variable that is changed in the experiment to test the effects on the dependent variable (Helmenstine, 2017). In another word, other measured variables have no effect on the tested variable. In this research, there are three independent variables Social Media Marketing, Brand Loyalty, and Electronic Word of Mouth used to carry out the independent experiment on its relationship with the dependent variable. On the other hand, the dependent variable is the variable being measured where any changes on the independent variable will affect the character of the dependent variable (Helmenstine, 2017). Purchase intention is the dependent variable used in this study to capture its effect based on the three changes in independent variables. There is a mediating variable in this study where it is caused by the independent variable and itself a cause to the dependent variable (Mustafe, 2015). The Brand Loyalty is the mediating variable for both Social Media Marketing, and Electronic Word of Mouth towards the Purchase Intention in this study.

3.9 Reliability Testing

Reliability referred to the degree which indicates consistent measurements (Hair et al, 2010; Leech, Barrett & Morgan, 2015; Sweet & Martin, 2012). Reliability testing Page 55 of 108

or measurement as according to Sekaran (2003), to indicate the extent to which it is without bias and consistently measured across time and at any different point in time. In other words, it is to ensure the responses do not vary across time periods so that any measurement at any point in time is reliable (Hair et al, 2010).

According to Sekaran (2003), and Hair et al (2010), the most popular interitem consistency reliability test is Cronbach’s coefficient alpha. Cronbach’s alpha is famous to test multipoint-scaled items with the higher the coefficients, the better the reliability. Alpha is widely used when the questionnaires involved several Likert-type items which summed to make a composite score (Leech et al, 2015). According to Leech and colleagues (2015), alpha referred to the average correlation of each item in the scale with others. The reliability coefficient of 0.70 and higher is considered acceptable in the social science research (Hair et al, 2010; UCLA, n.d.). According to Andale (2014a), alpha is sensitive to the number of items in a test. Andale (2014a) argued that if the alpha value is high, it may indicate there are redundant questions. Similarly, a low value of alpha might suggest insufficient questions on the test.

Table 2: Cronbach's Alpha Coefficient Range Alpha Coefficient Range

Indication

0.0 – 0.6

Poor

0.6 – 0.7

Moderate

0.7 – 0.8

Good

0.8 – 0.9

Very Good

0.9 – 1.0

Excellent

Page 56 of 108

3.10 Statistical Testing

The objective of statistics is to test the hypothesis. It is a method for the researcher to test the results of the survey to see if the results are meaningful or not. According to Andale (2017c), hypothesis testing is to determine the results of the experiment or survey could have happened by chance.

3.10.1 Linear Regression

Linear regression analysis is a model that testing the relationship between two variables. In other words, linear regression analysis is an attempt to explain that the increase of independent variable is associated with a consistent and constant increase or decrease in the dependent variable (Sweet & Martin, 2012).

3.10.2 Multiple Regression Analysis

Multiple regression analysis is useful in testing the relationship between the sets of variables. It involved multiple predictors “x” compared to the simple linear regression that uses single predictor for each dependent “y” variable (Andale, 2014b). In other words, simple linear regression usually is not enough to account all the factors that influence an outcome. And in this study, the independent variables Social Media Marketing, Brand Loyalty, and Electronic Word of Mouth are tested against the dependent variable’s Purchase Intention. Hence it is useful to test it with multiple regression analysis.

Page 57 of 108

3.10.3 Mediation Effect Analysis

Mediation effect analysis is a test that involved mediating variable. On a direct effect relationship, independent variable has direct influence over the dependent variable. However, mediation effect is mediated by a mediator that intervening the direct relationship between the independent variable and dependent variable. In other words, independent variable no longer affects dependent variable after the mediator has been in controlled (StatisticsSolutions, n.d.). PROCESS macro is used in this study to analyse the mediation effect in SPSS and written by Hayes (Hayes, 2016). In the PROCESS macro, Model 4 will be applied for the analysis due to the relationship between independent variable to mediating variable, and to the dependent variable.

3.11 Summary

In this chapter overall described the methodology applied to conduct the research and how data are collected for analysis in Chapter 4. The research design applied Saunders’s research ‘onion’ where every layer is discussed to recognise the best design to adapt for this study. Research paradigm namely ontology and epistemology described how knowledge is regarded and the relation to the knowledge. In this study, positivist philosophies approach was applied with deductive research approach. On the research methodological choices, the quantitative method was applied where the survey is incorporated to collect sample data from the large population. This chapter discussed as well on the construct measurements and its reliability testing to ensure the consistent measurements. The statistical analysis methods were discussed as well to incorporate for later analysis in Chapter 4.

Page 58 of 108

CHAPTER 4

RESULTS

4.1 Introduction

This chapter outline the collected data from survey and analyses are done via statistical software SPSS. The methodology used in this research to collect the data for later analysis through statistical software SPSS to analyse the results against the hypotheses set earlier. A total of 201 set of data were collected for this survey. 1 set of the data was incomplete and were discarded.

Page 59 of 108

4.2 Respondent Demographics

From the 200 sets of data, there are 2 respondents decided to be anonymous on their gender. On the 198 sets of data, the representation of the respondents’ demographic profiles is tabulated in Table 3 as an overview. Table 3: Respondents Demographic Information Respondent, N = 200 Variable

Category

Gender

Male

88

44.4

44.4

Female

110

55.6

100

No response

2

0

0

Below 20

1

0.5

0.5

20 - 29

61

30.5

31

30 - 39

93

46.5

77.5

40 - 49

22

11

88.5

50 and above

23

11.5

100

Public sector

12

6

6

Private sector

133

66.5

72.5

Self-employed

29

14.5

87

Student

10

5

92

Retired / homemaker

16

8

100

1 - 5 years

5

2.5

2.5

6 - 10 years

25

12.5

15

11 - 15 years

55

27.5

42.5

16 years and above

115

57.5

100

All the time

170

85

85

4 - 5 times a week

15

7.5

92.5

1 - 2 times a week

9

4.5

97

A few times a month

6

3

100

Age

Occupation

Internet Use

Social Media Use

Frequency

Valid Percent

Cumulative Percent

Based on Figure 4, a total of 88 male responded to this survey and represented a valid of 44.4% of the total. There are 110 female respondents participated in this survey with a valid of 55.6% represented the largest group.

Page 60 of 108

Figure 4: Gender Demographic Information

Gender

Male, 88 Female, 110

Figure 5 provides an overview of the age group demographic information of the collected samples. The largest representation is those within the age of 30 to 39 which accounted 46%, almost half of the sample size. Then follow by the generation of 20 to 29 at 30%. Both the groups aged 40 to 49, and 50 and above accounted for 23%. Interestingly, the sample below the age of 20 cumulated less than 1%.

Page 61 of 108

Figure 5: Age Demographic Information

Age Groups Below 20 1% 40 - 49 11% 50 and above 12%

30 - 39 46%

20 - 29 30%

From Figure 6, the occupation variable shows that over 133 respondents are working in the private sector, while only 12 respondents from the public sector. The self-employed group represented the second largest occupation at 29 over 200 respondents. In the meantime, the retired/homemaker, and students accounted 16 and 10 each.

Page 62 of 108

Figure 6: Occupations Demographic Information

Based on the plot graph in Figure 7, more than 57% of the respondents had used internet for 16 years and above since the introduction of the internet back then. On the second placing are the respondents had used internet for 11 to 15 years. The respondents below 10 years of internet utilisation accounted for 30 respondents only.

Page 63 of 108

Figure 7: Years of Internet Utilisation

Years of Internet Utilization 140 115

120 100 80 55

60 40 25 20 5 0 1 - 5 years

6 - 10 years

11 - 15 years

16 years and above

According to the sample collected and tabulated in Figure 8, Malaysian are actively on social media platforms. This was proven that 170 respondents are active all the time on social media platforms. Only a small number of 30 respondents used less than 5 times a week.

Figure 8: Social Media Activeness

Social Media Activeness 180

170

160 140 120 100 80 60 40 15

20

9

6

1 - 2 times a week

A few times a month

0 All the time

4 - 5 times a week

Page 64 of 108

4.3 Descriptive Analyses

Figure 9: Popular Social Media Platforms in Malaysia

Popular Social Media Platforms in Malaysia 200

181

120

100

160

141

140

80

120

105

110

91

100

60

80 54

60

74

67

40 20

56

17

24

43

9

12

5 4

6 6

3 14

9 15

15 39

Reddit

Tumblr

Weibo

Snapchat

Pinterest

48

42

40

Soial Media Platform (Frequency)

Gender Distribution on Social Media Platforms

180

20

23 32

48

62

61

73

105

0

0 Twitter

Google+

Instagram

LinkedIn

YouTube

Facebook

Social Media Platforms

Male

Female

Total

From the above bar chart (Figure 9), social media platforms are very popular in Malaysia. The chart shows that Facebook is the most popular social media platform in Malaysia, followed by YouTube, LinkedIn, Instagram, and the others. Further, analyse the chart, female respondents are the most active on social media platform. Among the social media platform popularity, Facebook ranked the first on the popularity list. And there is a huge gap between both female and male respondents. The gap is significant on Facebook, followed by Instagram and Pinterest social media platforms. While other platforms are well known but not popular among the respondents in Malaysia.

Page 65 of 108

Figure 10: Active Age Groups on Social Media Platforms

Active Age Groups on Social Media Platforms 87 64 41 31

30

2426 17 16 107 6 54 56 42 1 0 00 0 10 0 00 0 20 0 33 0

Below 20

20-29

30-39

54 44

99 0

40-49

52

48 44

0

11 6

51

58

21 15

1114 0

0

50 & above

The popularity of social media platforms had played a significant role among the respondents as in Figure 9. While in Figure 10, it shows which category of respondents is the most active on social media platforms. The age group between 30 to 39 are the most active group on social media platforms, especially on Facebook. It also followed closely by the 20 to 29 age group’s respondents. Those who are aged between 40 and above did not imply that they are highly active on the social media platforms.

Page 66 of 108

Figure 11: Brand Consciousness Among Gender

Brand Consciousness Among Gender 3 No 7

42 Not Really 25

65 Yes 56

Female

Male

Based on Figure 11, brand consciousness is very high among the respondent group. This indicates that they do have several brands that they preferred. No doubt the female is the highest respondent group in this survey, and they are highly brand conscious where they do have few brands they love. While the second group of respondents buy what they like or needed and did not particularly focus on the few brands only. Only a very small group of respondents do not care about brands consciousness.

4.4 Reliability Analysis

Reliability referred to the degree which indicates consistent measurements (Hair et al, 2010; Leech et al, 2015; Sweet & Martin, 2012). Reliability testing or measurement as according to Sekaran (2003), to indicate the extent to which it is without bias and consistently measured across time and at any different point in time. According to Leech and colleagues (2015), alpha referred to the average

Page 67 of 108

correlation of each item in the scale with others. The reliability coefficient of 0.70 and higher is considered acceptable in the social science research (Hair et al, 2010; UCLA, n.d.).

From Table 4, the reliability test was conducted on each of the components for its consistency. The result for Social Media Marketing (SMM) was 0.707 for 5 questions. While Brand Loyalty was reliable at 0.794 with 5 questions as well. eWOM achieved 0.744 for 4 questions and lastly, Purchase Intention achieved 0.832 with 3 questions. From the results, only Purchase Intention component exceeded the Good range, between 0.7 to 0.8 and achieved a Very Good rating. The other components satisfied the Cronbach’s Alpha reliability test at the Good range. As the alpha is sensitive to the number of items in a test, all the questions in each component did not face the issues of redundant questions or insufficient questions.

Table 4: Reliability Analysis Component Social Media Marketing Brand Loyalty EWM PI

Questions SMM1, SMM2, SMM3, SMM4, SMM5 BL1, BL2, BL3, BL4, BL5 EWM1, EWM2, EWM3, EWM4 PI1, PI2, PI3

Page 68 of 108

Cronbach’s Alpha 0.707 0.796 0.744 0.832

4.5 Linear Regression Analysis

Simple regression method was applied to examine H3, H4, and H5 hypotheses towards the dependent variable which is purchase intention from social media marketing, brand loyalty, and eWOM. Table 5: Model Summary (H3) Model Summary

Model

R

R Square

.363a

1

Adjusted R

Std. Error of the

Square

Estimate

.132

.128

.70221

a. Predictors: (Constant), SMM_5

The R value represents the simple correlation and is 0.363. The R2 value indicates how much of the total variation in the dependent variable, Purchase Intention, can be explained by the independent variable, Social Media Marketing. In this case, 13.2% can be explained. And 86.8% of the total variation in the dependent variable will be addressed by other variables and unknown factors.

Table 6: ANOVA (H3) ANOVAa Model 1

Sum of Squares

df

Mean Square

Regression

14.862

1

14.862

Residual

97.633

198

.493

112.495

199

Total

F 30.140

Sig. .000b

a. Dependent Variable: PurchaseIntention b. Predictors: (Constant), SMM_5

The ANOVA table explained the linear regression model with F (1, 198) = 30.14 and is significant with p-value < 0.001. This indicates that the predictor significantly predicts the purchase intention.

Page 69 of 108

Table 7: Coefficients (H3) Coefficientsa Standardised Unstandardised Coefficients Model 1

B (Constant) SMM_5

Coefficients

Std. Error 2.017

.319

.464

.085

Beta

t

.363

Sig.

6.328

.000

5.490

.000

a. Dependent Variable: PurchaseIntention

The unstandardised coefficient (B) for predicting purchase intention from social media marketing variable is 0.464; the standardised coefficient (β) is 0.363; the significance level (Sig.) is .000. This indicates that every one-percentage increase in the social media marketing, there is a predicted increase in the purchase intention of 2.481. From the Coefficients table, the T value of 5.49 (sig=0.000) shows that social media marketing has a positive influence on purchase intentions at 99% confidence level.

The general form of the equation to predict purchase intention from social media marketing is

Purchase Intention

= 2.017 + 0.464 Social Media Marketing

From the above equation, purchase intention (dependent variable) is equal to 2.017 when social media marketing is zero. Therefore, the dependent variable is expected to increase by 0.464 units if one unit is increased in social media marketing. The equation is as following

Purchase Intentions

= 2.017 + 0.464(1) = 2.481

Page 70 of 108

Table 8: Model Summary (H4) Model Summary

Model

R

Std. Error of the

Square

Estimate

R Square

.473a

1

Adjusted R

.224

.220

.66400

a. Predictors: (Constant), BrandLoyalty

The R value represents the simple correlation is 0.473. The R2 = 0.224 or 22.4% of the total variation in the dependent variable, Purchase Intention, can be explained by the independent variable, brand loyalty. While the remaining could be explained by other factors.

Table 9: ANOVA (H4) ANOVAa Model 1

Sum of Squares

df

Mean Square

Regression

25.199

1

25.199

Residual

87.296

198

.441

112.495

199

Total

F 57.154

Sig. .000b

a. Dependent Variable: PurchaseIntention b. Predictors: (Constant), BrandLoyalty

The ANOVA table shows that F (1, 198) = 57.154 is significant with p-value < 0001. This indicates that the brand loyalty significantly predicts purchase intention.

Table 10: Coefficients (H4) Coefficientsa Standardised Unstandardised Coefficients Model 1

B (Constant) BrandLoyalty

Std. Error 1.841

.256

.521

.069

a. Dependent Variable: PurchaseIntention

Page 71 of 108

Coefficients Beta

t

.473

Sig.

7.188

.000

7.560

.000

The unstandardised coefficient (B) for predicting purchase intention from brand loyalty variable is 0.521; the standardised coefficient (β) is 0.473; the significance level (Sig.) is .000. This indicates that every one-percentage increase in the social media marketing, there is a predicted increase in the purchase intention of 2.362. From the Coefficients table, the T value of 7.56 (sig=0.000) shows that brand loyalty has a positive influence on purchase intentions at 99% confidence level.

The general form of the equation to predict purchase intention from brand loyalty is

Purchase Intention

= 1.841 + 0.521 Brand Loyalty

From the above equation, purchase intention (dependent variable) is equal to 1.841 when brand loyalty is zero. Therefore, the dependent variable is expected to increase by 0.521 units if one unit is increased in brand loyalty. The equation is as following

Purchase Intentions

= 1.841 + 0.521(1) = 2.362

Table 11: Model Summary (H5) Model Summary

Model 1

R .658a

R Square .433

Adjusted R

Std. Error of the

Square

Estimate .430

.56770

a. Predictors: (Constant), EWOM

The R value represents the simple correlation and is 0.658. The R2 value indicates how much of the total variation in the dependent variable, Purchase Intention, can be explained by the independent variable, eWOM. In this case, 43.3% can be explained and the remaining can be explained by other factors. Page 72 of 108

Table 12: ANOVA (H5) ANOVAa Model 1

Sum of Squares

df

Mean Square

Regression

48.684

1

48.684

Residual

63.811

198

.322

112.495

199

Total

F 151.062

Sig. .000b

a. Dependent Variable: PurchaseIntention b. Predictors: (Constant), EWOM

The ANOVA table shows that F (1, 198) = 151.062 and is significant with p-value < 0.001. This indicates that the predictor significantly predicts purchase intention.

Table 13: Coefficients (H5) Coefficientsa Standardised Unstandardised Coefficients Model 1

B (Constant) EWOM

Std. Error 1.088

.220

.700

.057

Coefficients Beta

t

.658

Sig.

4.951

.000

12.291

.000

a. Dependent Variable: PurchaseIntention

The unstandardised coefficient (B) for predicting purchase intention from brand loyalty variable is 0.700; the standardised coefficient (β) is 0.658; the significance level (Sig.) is .000. This indicates that every one-percentage increase in the social media marketing, there is a predicted increase in the purchase intention of 1.788. From the Coefficients table, the T value of 12.291 (sig=0.000) shows that eWOM has a positive influence on purchase intentions at 99% confidence level.

The general form of the equation to predict purchase intention from eWOM is

Purchase Intention

= 1.088 + 0.700 eWOM

Page 73 of 108

From the above equation, purchase intention (dependent variable) is equal to 1.088 when eWOM is zero. Therefore, the dependent variable is expected to increase by 0.700 units if one unit is increased in eWOM. The equation is as following

Purchase Intentions

= 1.088 + 0.700(1) = 1.788

All three hypotheses have significant impact on purchase intentions where it was significant at p-value